2026· Advances in Artificial Intelligence and Machine Learning· 0 citations· 37 references
TL;DR
This paper presents a supervised attention-based deep learning approach for separating single-channel audio signals that establishes a low complexity, scalable, and interpretable solution to realistic problems in source separation.
Abstract
Audio source separation is the process of separating a mixed audio signal into its individual sources. This task remains difficult because of significant overlap in time and frequency. The presence of noise and fluctuating signal characteristics further deteriorate the task. Traditional signal processing methods rely on strong statistical assumptions, which often limit their effectiveness in real-world acoustic signal processing. To overcome some of these challenges, this paper presents a supervised attention-based deep learning approach for separating single-channel audio signals. This approach directly predicts the individual source magnitude spectrograms without employing traditional time-frequency masking methods. The approach is based on fixed-length segments of audio signals, which are represented at a constant rate and mapped to magnitude spectrograms using short-time Fourier transform (STFT). A structured preprocessing employed in the algorithm ensure normalisation, cropping, and temporal alignment, maintains constant input data size during training and testing phases. The separation approach employs a DenseNet-121 network that uses hierarchical spectral feature extraction. Thereby enabling the effective extraction of complex harmonic and formant characteristics. To predict time-frequency regions containing salient perceptual significance, circular spatial attention mechanism has been employed for enabling the network to concentrate on significant acoustic elements. The temporal characteristics and contextual parameters from spectrum frames are captured using a bidirectional long shortterm memory (Bi-LSTM) network and multi-head self-attention. The combined feature representations are processed through fully connected layers of the DenseNet-121 network to estimate the magnitude spectrograms of the constituent sources. The setup allows for endto-end learning of nonlinear source mappings. The network is trained using scale-invariant signal-to-distortion ratio (SI-SDR) as loss function and the trained network is evaluated using objective metrics, such as short-time objective intelligibility (STOI) and perceptual evaluation of speech quality (PESQ). The experimental results demonstrate satisfactory performance of separation, intelligibility, and perceptual quality across different mixing ratios. The approach establishes a low complexity, scalable, and interpretable solution to realistic problems in source separation.
It has long been recognized that the accurate classification of audio signals is a major challenge in the field of speech and signal processing research. Applications range from speaker recognition, speech emotion detection, and environmental sound categorization. This is especially challenging due to the dynamic natur...
Jayapadmini Kanchan, K. S. Pradyuthi, C. V. Swaraj et al.· International Conference on...· 0 citations
Sound source localization remains vulnerable in challenging acoustic environments characterized by reverberation and low signal-to-noise ratios. To address this issue, this paper presents spatial self-attention-driven feature re-weighting within a convolutional neural network (SSA-CNN). The re-weighting mechanism combi...
Ye Tian, Li-Qin Tian, Yan-Ru Feng· IEEE Access· 0 citations
A novel framework that leverages log-Mel spectrogram representations to capture informative time–frequency characteristics of audio signals and is further enhanced by a Dual-Attention Feature Fusion mechanism to improve discriminative capability at both spatial and channel levels is proposed.
Kusnaeni Kusnaeni, Muhaimin Hading, Z. Zaitun et al.· Journal of Mathematics, Comp...· 0 citations
Pretrained audio spectrogram transformers are widely used as backbones of machine listening systems for sound event classification. However, they are typically trained using spectrograms with a fixed Short-Time Fourier Transform (STFT) configuration, implicitly binding the model to the time-frequency resolution seen du...
Julia Wilkins, Wei-Ching Lin, Abinaya Kumar et al.· 0 citations
Separating overlapping talkers and suppressing non-stationary noise are coupled problems, yet conventional pipelines solve them in isolation, so separation artifacts propagate untouched into the denoising stage. We therefore build a time–frequency joint separation–denoising network: a learnable encoder, a multi-scale...
Rong-Wen Xie, Zeng-Jie Tao, Lin Lei et al.· Scientific Reports· 0 citations
Audio classification involves analyzing sound recordings and assigning them to specific categories. However, it faces several challenges, including unwanted noise, irregular phoneme articulation and scarcity or abundance of data due to reduced classification accuracy. This research proposed Wavelet Autoencoder and Conv...
Joylin Priya Pinto, R. Balasubramani· International Conference on...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.